Skip to content

Real-Time EEG Streaming

Coming Soon

Real-time EEG streaming via Lab Streaming Layer (LSL) is planned for a future release. The API is designed but not yet implemented.

Installation

pip install aime-loc[eeg,realtime]

This adds pylsl for Lab Streaming Layer protocol support.

Planned API

The real-time API will allow continuous TC scoring from a live EEG stream:

from aime_loc import LOC
from aime_loc.eeg import EEG

loc = LOC()
eeg = EEG(loc)

# Connect to LSL stream
stream = eeg.stream(source="lsl", buffer_sec=2.0)

# Real-time scoring loop
for epoch in stream:
    profile = eeg.score(epoch)
    print(f"TC: {profile.tc_score:.1f}%")

Architecture

EEG Device → LSL Stream → Buffer (2s) → PSD → WebSocket API → TC Score
                                                    ↑
                                            Server-side scoring
                                            (<200ms latency)

Key design:

  • Buffer: Collects 2-second windows from LSL stream
  • Local PSD: Welch PSD computed locally (same as offline)
  • WebSocket: Low-latency scoring via WS /v1/eeg/stream endpoint
  • Latency: Target <200ms end-to-end (PSD computation + API round-trip)

Use Cases

  • Neurofeedback: Real-time TC feedback during meditation or cognitive training
  • BCI Research: Cognitive state monitoring for brain-computer interfaces
  • Clinical Monitoring: Continuous consciousness assessment

Workaround: Near-Real-Time with Offline Pipeline

Until the real-time API is available, you can approximate near-real-time scoring using short recordings:

import time

while True:
    # Record 5 seconds of data (via your device's SDK)
    data = record_from_device(duration=5.0, sfreq=256)

    # Score immediately
    recording = eeg.load(data, sfreq=256, device="muse")
    recording.preprocess()
    epochs = recording.extract_epochs(duration=2.0)
    profile = eeg.score(epochs)

    print(f"TC: {profile.tc_score:.1f}%")
    time.sleep(5)

Next Steps